The hack is incredibly simple. First, an attacker signs up for a social login account at LinkedIn or Amazon using the victim’s email account. (Stipulation: the victim must not already have an account with that identity provider.) The victim then receives an email to verify the new account.

Here’s the catch: email verification is not needed to pull of the second phase of the hack. The attacker instead heads over to a website that supports logging in using a social account and selects “Sign In with LinkedIn” or “Log In with Amazon” or whatever relevant choice. If the attacker selects the just-created but not verified account and the victim already has a profile on the site in question—both associated with the same email address—vulnerable sites will authenticate the attacker, enabling him or her to assume the victim’s identity.

And that’s where the real trouble starts. An attacker could masquerade as a public-company executive on Nasdaq.com and comment on stocks, impacting the company’s stock performance. An attacker could post malicious links on the site under the assumed identity, subjecting anyone inquisitive enough to click to a phishing attack that allows the hacker to obtain sensitive information (that could quickly cascade into many more compromises).

“That’s a huge gaping security hole,” says Marla Hay, a senior product manager at Janrain, a company that connects websites to identity providers with social logins. (Some of the vulnerable sites—including Nasdaq.com and Slashdot.com—use the Janrain’s services.) “This is a super easy attack to make,” she adds.

Jamie Beckland, vice president of marketing at Janrain, says the onus falls on third-party websites to ensure that they’re implementing social logins securely. Even so, when Beckland’s team learned of the vulnerability through a VentureBeat article, it reached out to its clients within an hour to help them fix the flaw.

Hay recommends that companies with websites incorporating social login—so-called relying websites—take several measures. First, they should make sure to set up a field requesting a verified email address—not just any old unverified email address. Second, they should bar users from authenticating without first having verified email addresses. Lastly, they should consider accepting only identity providers that require users to verify email addresses before enabling validation through social login.

In a blog post, Or Peles, an IBM security researcher who co-authored the whitepaper describing the attack (which has been dubbed “SpoofedMe”), writes: “While fixing the identity provider vulnerability would be enough for this attack to be blocked…it is important for websites that are vulnerable to fix the website design problem because it may expose their users to similar attacks.”

Below, a video demonstration of the attack compiled by the researchers:

[youtube https://www.youtube.com/watch?v=kC0s3S00Dmk?rel=0&showinfo=0]

Next, read: This company wants to end web anonymity